[OLPC Brasil] Why call tools "activities"?

[Albert Cahalan]

On 4/5/07, Don Hopkins <dhopkins at donhopkins.com> wrote:
> Albert Cahalan wrote:

> > Componentizing the world is a failed dream.
> I strongly disagree with that.
>
> Just because C++ is totally fucked, CORBA is a sick joke, and Apple gave
> up on HyperCard and OpenDoc and ScriptX, and regressed to using the
> quaint circa-1989 technology of NeXT Step and Objective C, doesn't mean
> components are a failure -- it just means that Apple and other companies
> have failed to deliver on their own promises. Other important component
> based systems like Smalltalk, Python, SWIG, Java, Eclipse, web services,
> ReST, and even Microsoft Window's OLE/ActiveX are quite successful.

Most of those are failures. Python and Java don't exactly count.

You got one though: OLE/ActiveX. This is the very successful
technology that lets LARGE MONOLITHIC APPS do various
embedding tricks. Alternately you may prefer to argue that the
various Microsoft Office apps are modularized components that
represent good examples to follow; in that case you like apps.

In any case, "app" does not exclude modularization and
componentization. It doesn't even exclude toy and academic
languages. It does exclude "eating in the cafeteria" though, and
it avoids the confusion of performing activities with activities.

> Python and Sugar is all about modularization and componentization, and
> if don't think that's a worthy goal, or believe it's destine for
> failure, then you should be working on something else more monolithic,
> like a big "application".

Heh. It's not a dirty word. There are numerous apps that are
excellent for kids. Some are way more usable than the stuff
shipping on the laptop. Some are even modular.

Note the spelling: "app" Nobody says "application" anymore.

Apps beat buzzwords any day.